The purification and characterization of human hematopoietic stem cell (HSC) subsets has been facilitated by the availability of a variety of informative in vitro culture systems, and by the use of xenograft models of human hematopoiesis in mice and sheep. Although not "ideal", these xenograft systems can serve as surrogates for the evaluation of the in vivo engraftment potential of human HSC, the ultimate test by which the in vivo potential of these cells can be assessed. The investigators took advantage of fetal immunologic immaturity and developing "homing" niches in the fetal bone marrow to achieve human HSC engraftment in sheep without marrow conditioning or the use of cytokines. In the human/sheep xenograft model, human HSC: 1) colonizes the bone marrow; 2) persists for many years; and 3) is capable of multilineage differentiation. A unique feature of the sheep model is its large size, which permits repeated evaluation of human cell activity in the same animal over several years. Long-term observations have shown that in this model, human cells retain their ability to respond to human cytokines and to engraft in secondary recipients. More importantly, initial studies indicate that the sheep model can discriminate between the different populations of human HSC and exhibits a degree of sensitivity that suggest this in utero approach may serve as a biologically relevant model for the identification, characterization, and study of the in vivo potential of human HSC from available sources. The overall aim of this renewal application is to further improve the model so that it can serve as a reliable, sensitive and specific, and relatively rapid test system for candidate human HSC. To achieve this aim, the applicants plan to 1) improve human HSC engraftment in sheep by the use of co-transplantation of autologous stroma, and, if necessary, a multiple transplantation approach; 2) clearly establish the specificity of the model by evaluating its ability to distinguish between short-term and long-term engrafting human HSC subsets using observations in primary and secondary transplant recipients; 3) establish the sensitivity of the model by careful dose-response studies in the presence or absence of autologous stroma; and 4) develop a more rapid version of the test system that reliably differentiates between the different functional subsets of human HSC. They will also continue to apply the model to the assessment of the in vivo engrafting potential of different preparations of candidate human HSC. It is hoped that these experiments will help establish an in vivo assay system for human HSC that can reliably predict the clinical usefulness of different phenotypically/functionally defined human HSC subsets.